Shell connects its engineers, and then gets out of the way

Oil refineries are a 24 hours a day, 7 days a week operation. An important step in the process of refining is Fluid Catalytic Cracking. This is done with a machine referred to commonly as the CAT cracker. During this process, crude oil is converted into more valuable derivaties, ranging from heavier oils like diesel, to lighter ones like gasoline, kerosine, and gasses.

The CAT cracker has an important role for the yield of the refinery operations. Because it can be used to produce multiple kinds of oil derivatives, the CAT cracker is the oil company’s tool to better cater to fluctuations in market demand for derivative products. With a CAT cracker in place, an oil plant can always select to produce for the product that fetches the higher price in the market.

Essentially the Cat Cracker works like a distillation process. It heats up crude oil to a temperature between 315, and 400 degrees Celsius under pressure. Then by using a powdered catalyst, a reactor breaks up the long chained crude oil molecules into smaller, lighter components, which are the derivatives.

Once every 4 years the CAT cracker needs to be shutdown for maintenance. The shutdown process is a key influence on the yield of an oil refinery. This maintenance puts operations down for about 7 weeks on average. Of about 640 oil refineries world wide, an estimated 400 use CAT crackers. Given that an average-sized refinery makes around 2 million euro per day from product that’s produced by a CAT cracker, the global costs of refinery downtime is 10 billion dollars a year. This is a significant amount for an industry that is under a lot of volatility because of fluctuating crude oil prices. Every day saved on maintenance, means less risk to profitability.

Although it seems like a standard engineering procedure, the shutdown process is filled with uncertainties. For one, the CAT cracker is heated when it’s in operation, and needs to cool down before a maintenance crew can access it. Then there is the question of how extensive the maintenance needs to be, as the precise estimation of maintenance requirement can only be done upon visual inspection, after the reactor is actually down. Only then will you see how long the downtime is really going to be.

People, and their ability to coordinate, and collaborate are of the essence. Maintenance teams put meticulous planning into each shutdown to keep downtime to a minimum.

To improve the shutdown process, maintenance teams from various plant locations put their heads together every few years to share notes, in so-called Fat Rat workshops. Former head of refinery maintenance, Hans Peter Doorneweert, organised a couple of these workshops, and explains how they go down.

There is little structure to the workshops. The formula for success is basically to bring operational people together from 3 or 4 different plant sites. I invite people specifically from operations, maintenance and technology. No higher management! The focus is on the actual maintenance steps, and you need to put the brains together that are working on the ground.

In a 4-day workshop, participants are flown into a globally well-connected location, like Singapore. In the workshop, the shutdown process is dissected. We start at the very beginning with how they prepare for a shutdown, then move into the maintenance analysis, and the maintenance itself, to starting up operations again. In each step, anyone can interject. The idea is that everybody shares notes.

The operational workers who participate, are people that have worked on a single site for years, sometimes even 20 or more. They know all about how things go down at their plant. Yet, they only have exposure to their own location. They don’t ever travel. These workshops are often their first trip abroad for work ever. Some don’t even have company email addresses.


Maintenance work in the refinery during shutdown .

Because the engineers only have exposure to their own location, there is an abundance of experience to share between the engineers

You’d be surprised by the cultural differences between plants, and the way crews go about things. During the Fat Rat workshop, the lessons flush up quickly.

In one workshop a big win was in time saved to actually shut down operations. I had a team from the United Kingdom explain their preparation. If they were planning to start maintenance on a Monday, then they would begin on the Friday before, to gradually phase down operations, step by step. After explaining each phase of their process, another participant from the United States interjected, and said: “Why this lengthy business? I just push the button, and shut the reactor down instantly. If something breaks because of it, it’s fine. You’re performing maintenance right?!” That tip alone saved the plant in the UK 2 days of operations.

Another time, an engineer from the United States shared his practice for prioritising his maintenance process. He said: “The air grid always determines the largest part of your actual maintenance time, so I want to get to that first (The air grid blows air into the reactor, and is always under high temperature, and pressure stress). Usually, I’ll have an assessment of the status of the air grid within 24 hours after shutdown”. “That’s impossible!” all the rest said; “The reactor is still too hot to handle then”. And so the engineer form the US shared his tricks for cooling down the air grid faster.

And there was this story about Frenchman who found a way to speed up the process of burning off coke (carbon residue left after the incomplete combustion of the crude oil) from the reactor. To burn off the reactor, it is fed by compressed air of 6 Bar, which is a standard industrial air compressor. It takes about 4 or 5 days to burn off all the coke in the reactor.

This French guy said: “We take an air compressor of 12 Bar because you know: if you double the air pressure you half the burning time. His colleague from Norway responded: “Do those kind of air compressors exist!?”
The Frenchman said: “Of course they exist, we hire these specially from a firm in Rotterdam, I will give you the address.”

This 2 minute exchange paid for the whole workshop.

Overall, what seems like small operational tweaks, together add up to a considerable saving. Even though only 3 plants participated, on average they saved around 2 days of downtime. A single 4-day workshop results in 4 million euro savings per location.

The savings might be even more. After the workshops people stay in touch, and ask each other for advice. We share the insights from the workshops as a report with higher management as well, all around the world. Also we handed out a yearly trophy for the shortest shutdown times! Who knows who might be reading into that, and applying some the insights into their shutdown protocol.

In this case of the Fat Rat workshop, you see that you don’t need external expertise or management intervention to achieve concrete learning outcomes. It’s a matter of bringing together right mix of people, keeping conversation focussed, and its structure loose.

The loose workshop setup provides enough room for participants to ask questions, and share insights. There's no real program. The focus is on sharing the maintenance process. This setup allows participants to diagnose each other's situation, and come up with solutions.

By bringing in the right mix of participant, the Fat Rat workshop expands the networks of connections for the maintenance crews, who meet colleagues that they would normally not connect with.

If the company would invest more in connectivity, they could create a setting where maintenance crews from all over the world could connect with each other, and benchmark their performance, taking the effects of the workshop into a continuous learning, and adoption mode.